An Observational Study of the Causes, Management, and Outcomes of Community-acquired Sepsis and Severe Sepsis in Southeast Asia
Bacterial and viral infectious diseases are still the leading cause of death in Southeast Asia, comparable to injuries, cardiovascular diseases and other non-communicable diseases. Sepsis is defined as the body’s response to infectious diseases, including bacterial and viral causes. Sepsis should be considered as a medical emergency just like a heart attack or stroke because there is an interruption in the supply of oxygen and nutrients to the tissues, and emergency treatment is required. The definition of sepsis is broader than febrile illness as patients with severe infectious diseases may not present with fever, and infectious causes may be overlooked by physicians. On the other hand, it is not uncommon for patients who are diagnosed with sepsis on admission to later have a confirmed non-infectious diagnosis.
Recognition of sepsis as a specific syndrome is the most important key for improvements in diagnostic and treatment strategies for infectious diseases. The 1992 statement from ACCP/SCCM (American College of Chest Physicians and the Society of Critical Care Medicine) Conference introduced a definition of “systemic inflammatory response syndrome (SIRS)”. SIRS was considered when patients have more than one of the following clinical findings:
- Body temperature higher than 38oC or lower than 36oC
- Heart rate higher than 90/min
- Hyperventilation evidenced by respiratory rate higher than 20/min or PaCO2 lower than 32 mmHg
- White blood cell count higher than 12,000 cells/uL or lower than 4,000/uL
The European Society of Intensive Care Medicine (ESICM)/SCCM found a problem that 71% of clinicians and researchers cited no common definition of sepsis. In 2001, the criteria of SIRS were revised so that the updated criteria are useful to both clinicians caring for patients at bedside and to researchers designing observational studies. The revised criteria by SCCM/ESICM/ACCP/ATS/SIS were also to respond the sepsis research agenda to include the “standardization of research protocols”. The improvement was focused on an increase in sensitivity to identify most patients with infection, while minimally sacrifice inevitable specificity. The 2003 statement from SCCM/ESICM/ACCP/ATS/SIS Conference introduced the updated definition of sepsis and SIRS (Scheme 1). The use of this definition has been continuously recommended in 2004, 2008 and 2012 surviving sepsis campaign.
A recent study also confirmed that the 2003 SCCM/ESICM/ACCP/ATS/SIS sepsis definitions have higher sensitivity to include patients with clinically or microbiology proven infections compared to the 1992 ACCP/SCCM sepsis definition. Of 827 patients admitted to the ICU within a one year period, Weiss et al found that 16 of 282 patients with clinically or microbiologically proven infections did not meet the 1992 ACCP/SCCM sepsis definition, but all 282 patients with clinically or microbiologically proven infections met the 2003 SCCM/ESICM/ACCP/ATS/SIS sepsis definition. The 2003 SCCM/ESICM/ACCP/ATS/SIS sepsis definition also has a higher sensitivity to predict death. Therefore, we aim to use the current 2003 SCCM/ESICM/ACCP/ATS/SIS sepsis definition for inclusion criteria of sepsis patients (Scheme 1) in this sepsis study.
In addition, the recognition of sepsis cannot replace the diagnosis and management of specific clinical syndromes, such as acute diarrhea or acute encephalitis. However, in those patients, sepsis should be considered as an add-on to the diagnosis, and management of sepsis, in general, should not be overlooked. The recent Fluid Expansion as Supportive Therapy (FEAST) study highlights the need for a much greater understanding of this syndrome, its causes and management strategies in resource-limited settings.
Although infectious diseases are major causes of death in Southeast Asia, the epidemiology of its causes is poorly understood. This is largely because a number of tropical infectious diseases cannot be diagnosed by standard microbiology facilities. For example, most viral infections require specific PCR assays, and rickettsial diseases require an extra blood sample on admission and on convalescent phase for definite diagnosis. Understanding the epidemiology of the causes of sepsis is the corner stone to improve guidelines for diagnosis and management of sepsis specific to the area.
The Management of Severe sepsis in Asia’s Intensive Care units (MOSAIC) conducted a multicenter cohort study involving 16 countries in Asia. In 2009, results of this study were reported and indicated an overall hospital mortality rate from all causes of sepsis to be 44.5%. Similarly, a 2-year epidemiological sepsis study conducted at the Prince Songkla University Hospital in Thailand showed that the hospital and ICU mortality rate of severe sepsis and septic shock was 49.7% and 39.2% respectively and that an estimated 40.5% of the total ICU mortality cases of sepsis were felt to be caused by community-acquired disease. A retrospective review of data from 10 provincial hospitals in Northern Thailand compared to the national death registry between 2004 and 2010 revealed an incidence of community-acquired bacteremia (CAB) increasing from 16.7 to 38.1 per 100,000 people per year. Over that same time period the mortality rate associated with CAB increased from 6.9 to 13.7 per 100,000 people per year. Overall, of those patients admitted to the hospitals with CAB 37.5% died within 30 days of admission. These 3 studies serve to confirm the magnitude of the problem in Southeast Asia and the need to collect more empirical data. A review of the literature also confirms that there are very few population-based studies evaluating the incidence, causes and trends in CAB in low- and middle-income countries, an important omission since such information is required for strategic planning of available healthcare resources, together with disease prevention programs.
Current evidence suggests that a wide range of pathogens is responsible for sepsis and severe sepsis in Southeast Asia. An observational study conducted by Suttinont et al in 845 patients in 2006 showed that major causes of acute undifferentiated fever in Thai agricultural workers are leptospirosis (37%), scrub typhus (20%), influenza (11%), murine typhus (2.8%) or other bacterial infection (1.2%). Leptospirosis has also been reported as the major cause of fever in 13/137 in adults in Indonesia. Acute rickettsial infection was also identified as the cause of fever in 115/427 (27%) adults with negative blood cultures admitted to Mahosot Hospital in Vientiane, Laos, and in 9/137 in adults with acute undifferentiated fever in Semarang, Indonesia. Gram-negative aerobes accounted for 90% of causes of sepsis (67% Salmonella typhi, 10% Escherichia coli, 5% Klebsiella spp. and 3% Salmonella para-typhi B) in adults and children admitted to infectious disease hospitals in southern Vietnam. That study highlighted that over 70% of the Salmonella typhi isolated were multi-drug-resistant. Similarly, Salmonella enterica serotype Typhi predominates as the cause of bacteremia in Cambodia, while Burkholderia pseudomallei is a major cause of bacteremia in Thailand. However, estimates vary on the incidence of community-acquired sepsis caused by Burkholderia pseudomallei. A study in Southern Thailand of community-acquired severe sepsis caused by Burkholderia pseudomallei indicated a 3.8% incidence. A recent retrospective review of CAB cases in Northern Thailand between 2004 and 2010 indicated that Burkholderia pseudomallei was the sepsis causative agent in 20.2% of adults, 14.3% of children aged 1-15 years and 1.6% in infants less than one year of age.
In an 8-year-long review of blood culture isolates from patients presenting to the Emergency Department of the Hospital Universiti Sains Malaysia beginning from 2002, 55.2% were found to be caused by gram-negative aerobes. Of these, increased drug resistance was noted with less than 80% sensitivity to ampicillin, cotrimoxazole and ciprofloxacin. Other studies highlight that often the cause of sepsis is unknown and a retrospective review of community-acquired bacteremia studies between 1990 and 2010 in South and South-east Asia identified that pathogenic organisms were only isolated in 12% of adults and 7% of children. The paper argues that better detection would lead to better patient outcomes and improved patient management. Goal-directed therapy and appropriate administration of antibiotics within 6 hours was noted to improve outcomes in sepsis patients at Siriraj Hospital in Bangkok, Thailand.
The first aim of this proposal, therefore, is to determine the causes of severe acute infectious illnesses (presenting with sepsis and requiring hospitalization by attending physicians) at multiple study sites across Southeast Asia. This study will also establish a regional repository of samples with robust clinical data to facilitate future pathogen discovery programs and allow future development and assessment of novel diagnostic tests. As stated above, the causes and incidence of community-acquired sepsis and severe sepsis vary. Treatment is initiated based on a presumptive diagnosis taking into account the clinical presentation, initial laboratory data as well as the local guidelines for the management of sepsis and severe sepsis. In many cases a definitive confirmation of the causative agent is not known until much later if at all. While guidelines are available that provide guidance for the treatment of the common causes of sepsis, some causes require a different choice of antibiotics and anti-viral agents. For example:
- Burkholderia pseudomallei is intrinsically resistant to commonly used first-line antibiotics which are used to treat sepsis in Southeast Asia including penicillin, aminoglycosides and most cephalosporins. Regimens including ceftazidime or a beta-lactam antibiotic such as one of the carbapenems are also indicated.
- Listeria monocytogenes possesses a natural in-vitro resistance to older quinolones, fosfomycin, and expanded-spectrum cephalosporins. Treatment is typically based on a synergistic association of high doses of aminopenicillin (ampicillin or amoxicillin) and gentamicin. Although rifampin, vancomycin, linezolid, and carbapenems have been proposed as possible alternatives, trimethoprim is generally used in case of intolerance to beta-lactams.
- Sepsis caused by the Rickettsiae group of organisms is best treated by early initiation of specific antibiotics including doxycycline, chloramphenicol or a fluoroquinolone.
- Sepsis and encephalitis caused by Herpes simplex would warrant treatment with intravenous acyclovir.
When considering the prevalence of the different causes of sepsis that this study is likely to encounter at the study sites it is important to consider a sample size that will detect those causes that would lead to a different choice of antibiotics or anti-virals compared to standard therapy. An informal review of hospital records in Ho Chi Minh City in 2012 revealed that the prevalence of Listeria monocytogenes, Burkholderia pseudomallei, Herpes simplex and Rickettsiae organisms as the confirmed cause of community-acquired sepsis and severe sepsis was greater than 2% but less than 5%. Similarly, the recent retrospective study of CAB in Northern Thailand showed that if a 5% prevalence of the difference causes of disease was assumed then this would miss important causes including Burkholderia pseudomallei and Haemophilus influenzae in infants under one year of age (2% and 2.8% prevalence respectively), Streptococcus pneumonia in adults (2.5% prevalence), Group A Streptococcus in all age groups (2.2-3%), and extended spectrum beta lactamase (ESBL) Escherichia coli in adults (2.9% prevalence). These data suggest that a study powered to detect 5% prevalence of any particular cause would likely miss some important causes resulting in cause-appropriate treatment not being administered. This has obvious implications for patient survival, particularly when mortality rates are known to be high if appropriate treatment is not administered quickly. These data would also indicate that detecting a 2% prevalence of the different causes of sepsis would be more clinically appropriate and likely lead to higher survival rates for patients in this study.
Severe sepsis is defined as sepsis plus sepsis-induced organ dysfunction or tissue hypoperfusion. Around 20-40% of sepsis patients presenting to the hospitals develop severe sepsis, which is associated with a very high mortality rate. Apart from antibiotic and anti-viral treatment being an important component of sepsis survival, the overall comprehensive clinical management of the patient is critical, particularly for those who develop severe sepsis. In 2004, an international group of experts in the diagnosis and management of infectious diseases published the first internationally accepted guidelines that the bedside clinician could use to improve outcomes in severe sepsis (sepsis plus any acute organ dysfunction). This “surviving sepsis campaign (updated in 2008)” aims to increase awareness and improve the acute management in severe sepsis. Acute management of severe sepsis during the first 48 hours is very important to reduce mortality. The MOSAIC’s multinational cohort study reported that the overall compliance with the Surviving Sepsis Campaign’s resuscitation and management bundles for severe sepsis in adult patients in Asian countries was only 7.6% and 3.5% respectively. These numbers are much lower than those reported by other similar studies in Europe and the Americas.
Nonetheless, most of the interventions in the “surviving sepsis campaign” are based on facilities available in the developed world. It is very likely that management of sepsis and severe sepsis varies widely across Southeast Asia, in part due to limited resources.
The second aim of this proposal, therefore, is to observe current acute management (within the first 48 hours) of patients presenting with sepsis and severe sepsis to design practical interventions that could reduce the mortality in the future. This study will observe the routine practice in detection of severe sepsis among sepsis patients, and the overall management of patients with sepsis and severe sepsis at multiple study sites across Southeast Asia. This study will focus on observations of early and basic management, including fluid management, blood pressure control, timely administration of antibiotics, and source control, which is likely to be the most cost-effective intervention even in resource-constrained settings.
Outcome from patients presenting with sepsis and severe sepsis is dependent on many factors, including causes of infection, severity of infection, sepsis resuscitation, antimicrobial treatment and host genetic factors. There is a lack of studies showing genetic factors associated with outcomes in patients with sepsis and severe sepsis in developing countries. A recent retrospective study in 669 Thai patients with culture-proven melioidosis admitted to Sappasithiprasong Hospital shows that hypofunctional genetic variant TLR51174C>T is associated with reduced organ failure and improved survival in melioidosis. Understanding all predictive factors associated with inflammatory responses, organ failure, and mortality of patients with sepsis and severe sepsis, including genetic factors, would be very important to improve the management and outcome of this group of patients in the future. Therefore in this study we will ask adult subjects if they wish to provide one additional sample of blood to be stored for later genetic testing. This is optional and subjects can chose to not consent to providing this additional sample without affecting their eligibility to be part of the sepsis study.